Brain Inspired

Support the show to get full episodes, full archive, and join the Discord community.

The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.

Read more about our partnership.

Sign up for Brain Inspired email alerts to be notified every time a new Brain Inspired episode is released.

To explore more neuroscience news and perspectives, visit thetransmitter.org.

Michael Shadlen is a professor of neuroscience in the Department of Neuroscience at Columbia University, where he's the principle investigator of the Shadlen Lab. If you study the neural basis of decision making, you already know Shadlen's extensive research, because you are constantly referring to it if you're not already in his lab doing the work. The name Shadlen adorns many many papers relating the behavior and neural activity during decision-making to mathematical models in the drift diffusion family of models. That's not the only work he is known for,

As you may have gleaned from those little intro clips, Michael is with me today to discuss his account of what makes a thought conscious, in the hopes to inspire neuroscience research to eventually tackle the hard problem of consciousness - why and how we have subjective experience.

But Mike's account isn't an account of just consciousness. It's an account of nonconscious thought and conscious thought, and how thoughts go from non-conscious to conscious

His account is inspired by multiple sources and lines of reasoning.

Partly, Shadlen refers to philosophical accounts of cognition by people like Marleau-Ponty and James Gibson, appreciating the embodied and ecological aspects of cognition.

And much of his account derives from his own decades of research studying the neural basis of decision-making mostly using perceptual choice tasks where animals make eye movements to report their decisions.

So we discuss some of that, including what we continue to learn about neurobiological, neurophysiological, and anatomical details of brains, and the possibility of AI consciousness, given Shadlen's account.

Shadlen Lab.

Twitter: @shadlen.

Decision Making and Consciousness (Chapter in upcoming Principles of Neuroscience textbook).

Talk: Decision Making as a Model of thought

Read the transcript.

0:00 - Intro
7:05 - Overview of Mike's account
9:10 - Thought as interrogation
21:03 - Neurons and thoughts
27:05 - Why so many neurons?
36:21 - Evolution of Mike's thinking
39:48 - Marleau-Ponty, cognition, and meaning
44:54 - Naturalistic tasks
51:11 - Consciousness
58:01 - Martin Buber and relational consciousness
1:00:18 - Social and conscious phenomena correlated
1:04:17 - Function vs. nature of consciousness
1:06:05 - Did language evolve because of consciousness?
1:11:11 - Weak phenomenology and long-range feedback
1:22:02 - How does interrogation work in the brain?
1:26:18 - AI consciousness
1:35:49 - The hard problem of consciousness
1:39:34 - Meditation and flow

Support the show to get full episodes, full archive, and join the Discord community.

The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.

Read more about our partnership.

Sign up for Brain Inspired email alerts to be notified every time a new Brain Inspired episode is released.

To explore more neuroscience news and perspectives, visit thetransmitter.org.

Tomaso Poggio is the Eugene McDermott professor in the Department of Brain and Cognitive Sciences, an investigator at the McGovern Institute for Brain Research, a member of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and director of both the Center for Biological and Computational Learning at MIT and the Center for Brains, Minds, and Machines.

Tomaso believes we are in-between building and understanding useful AI That is, we are in between engineering and theory. He likens this stage to the period after Volta invented the battery and Maxwell developed the equations of electromagnetism. Tomaso has worked for decades on the theory and principles behind intelligence and learning in brains and machines. I first learned of him via his work with David Marr, in which they developed "Marr's levels" of analysis that frame explanation in terms of computation/function, algorithms, and implementation. Since then Tomaso has added "learning" as a crucial fourth level. I will refer to you his autobiography to learn more about the many influential people and projects he has worked with and on, the theorems he and others have proved to discover principles of intelligence, and his broader thoughts and reflections.

Right now, he is focused on the principles of compositional sparsity and genericity to explain how deep learning networks can (computationally) efficiently learn useful representations to solve tasks.

Lab website.

Tomaso's Autobiography

Related papers

Position: A Theory of Deep Learning Must Include Compositional Sparsity

The Levels of Understanding framework, revised

Blog post:

Poggio lab blog.

The Missing Foundations of Intelligence

Read the transcript.

0:00 - Intro
9:04 - Learning as the fourth level of Marr's levels
12:34 - Engineering then theory (Volta to Maxwell)
19:23 - Does AI need theory?
26:29 - Learning as the door to intelligence
38:30 - Learning in the brain vs backpropagation
40:45 - Compositional sparsity
49:57 - Math vs computer science
56:50 - Generalizability
1:04:41 - Sparse compositionality in brains?
1:07:33 - Theory vs experiment
1:09:46 - Who needs deep learning theory?
1:19:51 - Does theory really help? Patreon
1:28:54 - Outlook

Support the show to get full episodes, full archive, and join the Discord community.

The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.

Read more about our partnership.

Sign up for Brain Inspired email alerts to be notified every time a new Brain Inspired episode is released.

To explore more neuroscience news and perspectives, visit thetransmitter.org.

Alex is an associate professor of psychology at Vanderbilt University where he heads the Maier Lab. His work in neuroscience spans vision, visual perception, and cognition, studying the neurophysiology of cortical columns, and other related topics. Today, he is here to discuss where his focus has shifted over the past few years, the neuroscience of consciousness. I should say shifted back, since that was his original love, which you'll hear about.

I've known Alex since my own time at Vanderbilt, where I was a postdoc and he was a new faculty member, and I remember being impressed with him then. I was at a talk he gave - job talk or early talk - where it was immediately obvious how passionate and articulate he is about what he does, and I remember he even showed off some of his telescope photography - good pictures of the moon, I remember. Anyway, we always had fun interactions, even if sometimes it was a quick hello as he ran up stairs and down hallways to get wherever he was going, always in a hurry.

Today we discuss why Alex sees integration information theory as the most viable current prospect for explaining consciousness. That is mainly because IIT has developed a formalized mathematical account that hopes to do for consciousness what other math has done for physics, that is, give us what we know as laws of nature. So basically our discussion revolves around everything related to that, like philosophy of science, distinguishing mathematics from "the mathematical", some of the tools he is finding valuable, like category theory, and some of his work measuring the level of consciousness IIT says a whole soccer team has, not just the individuals that comprise the team.

Maier Lab

Astonishing Hypothesis (Alex's youtube channel)

Twitter:

Sensation and Perception textbook (in-the-making)

Related papers

Linking the Structure of Neuronal Mechanisms to the Structure of Qualia

Information integration and the latent consciousness of human groups

Neural mechanisms of predictive processing: a collaborative community experiment through the OpenScope program

Various things Alex mentioned:

“An Antiphilosophy of Mathematics,” Peter J. Freyd youtube video about "the mathematical".

David Kaiser's playlist on modern physics.

Here's a link to the Integrated Information Theory Wiki.

Read the transcript.

0:00 - Intro
4:27 - Discovering consciousness science
11:23 - Laws of perception
15:48 - Integrated information theory and mathematical formalism
23:54 - Theories of consciousness without math
28:18 - Computation metaphor
34:44 - Formalized mathematics is the way
36:56 - Category theory
41:42 - Structuralism
51:09 - The mathematical
54:33 - Metaphysics of the mathematical
59:52 - Yoneda Lemma
1:12:05 - What's real
1:26:22 - Measuring consciousness of a soccer team
1:35:03 - Assumptions and approximations of IIT
1:43:13 - Open science

Support the show to get full episodes, full archive, and join the Discord community.

The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.

Read more about our partnership.

Sign up for Brain Inspired email alerts to be notified every time a new Brain Inspired episode is released.

To explore more neuroscience news and perspectives, visit thetransmitter.org.

Can you look at all the synaptic connections of a brain, and tell me one nontrivial memory from the organism that has that brain? If so, you shall win the $100,000 prize from the Aspirational Neuroscience group.

I was recently invited for the second time to chair a panel of experts to discuss that question and all the issues around that question - how to decode a non-trivial memory from a static map of synaptic connectivity.

Before I play that recording, let me set the stage a bit more.

Aspirational Neuroscience is a community of neuroscientists run by Kenneth Hayworth, with the goal, from their website, to "balance aspirational thinking with respect to the long-term implications of a successful neuroscience with practical realism about our current state of ignorance and knowledge." One of those aspirations is to decoding things - memories, learned behaviors, and so on - from static connectomes. They hold satellite events at the SfN conference, and invite experts in connectomics from academia and from industry to share their thoughts and progress that might advance that goal.

In this panel discussion, we touch on multiple relevant topics. One question is what is the right experimental design or designs that would answer whether we are decoding memory - what is a benchmark in various model organisms, and for various theoretical frameworks? We discuss some of the obstacles in the way, both technologically and conceptually. Like the fact that proofreading connectome connections - manually verifying and editing them - is a giant bottleneck, or like the very definition of memory, what counts as a memory, let alone a "nontrivial" memory, and so on. And they take lots of questions from the audience as well.

I apologize the audio is not crystal clear in this recording. I did my best to clean it up, and I take full blame for not setting up my audio recorder to capture the best sound. So, if you are a listener, I'd encourage you to check out the video version, which also has subtitles throughout for when the language isn't clear.

Anyway, this is a fun and smart group of people, and I look forward to another one next year I hope.

The last time I did this was episode 180, BI 180, which I link to in the show notes. Before that I had on Ken Hayworth, whom I mentioned runs Aspirational Neuroscience, and Randal Koene, who is on the panel this time. They were on to talk about the future possibility of uploading minds to computers based on connectomes. That was episode 103.

Aspirational Neuroscience

Panel

Michał [email protected]

Research scientist (connectomics) with Google Research, automated neural tracing expert

Sven Dorkenwald

@sdorkenw.bsky.social

Research fellow at the Allen Institute, first-author on first full Drosophila connectome paper

Helene [email protected]

Group leader at Ernst Strungmann Institute, hippocampus connectome & EM expert

Andrew Payne

@andrewcpayne.bsky.social

Founder of E11 Bio, expansion microscopy & viral tracing expert

Randal Koene

Founder of the Carboncopies Foundation, computational neuroscientist dedicated to the problem of brain emulation.

Related episodes:

BI 103 Randal Koene and Ken Hayworth: The Road to Mind Uploading

BI 180 Panel Discussion: Long-term Memory Encoding and Connectome Decoding

Support the show to get full episodes, full archive, and join the Discord community.

The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.

Read more about our partnership.

Sign up for Brain Inspired email alerts to be notified every time a new Brain Inspired episode is released.

To explore more neuroscience news and perspectives, visit thetransmitter.org.

Tatiana Engel runs the Engel lab at Princeton University in the Princeton Neuroscience Institute. She's also part of the International Brain Laboratory, a massive across-lab, across-world, collaboration which you'll hear more about. My main impetus for inviting Tatiana was to talk about two projects she's been working on. One of those is connecting the functional dynamics of cognition with the connectivity of the underlying neural networks on which those dynamics unfold. We know the brain is high-dimensional - it has lots of interacting connections, we know the activity of those networks can often be described by lower-dimensional entities called manifolds, and Tatiana and her lab work to connect those two processes with something they call latent circuits. So you'll hear about that, you'll also hear about how the timescales of neurons across the brain are different but the same, why this is cool and surprising, and we discuss many topics around those main topics.

Engel Lab.

@engeltatiana.bsky.social.

International Brain Laboratory.

Related papers:

Latent circuit inference from heterogeneous neural responses during cognitive tasks

The dynamics and geometry of choice in the premotor cortex.

A unifying perspective on neural manifolds and circuits for cognition

Brain-wide organization of intrinsic timescales at single-neuron resolution

Single-unit activations confer inductive biases for emergent circuit solutions to cognitive tasks.

0:00 - Intro
3:03 - No central executive
5:01 - International brain lab
15:57 - Tatiana's background
24:49 - Dynamical systems
17:48 - Manifolds
33:10 - Latent task circuits
47:01 - Mixed selectivity
1:00:21 - Internal and external dynamics
1:03:47 - Modern vs classical modeling
1:14:30 - Intrinsic timescales
1:26:05 - Single trial dynamics
1:29:59 - Future of manifolds